Physical AI is moving into the real world.

But once a robot leaves the lab, Ali Agha says edge cases become the environment.

In this episode of Automated, recorded live at Automate in Chicago, Brian Heater speaks with Ali Agha, founder and CEO of FieldAI, about building autonomous robots that can operate safely in unpredictable, unstructured places.

Ali’s path to FieldAI runs through MIT, Qualcomm, NASA JPL, and DARPA. At Qualcomm, he worked to put autonomy on the low-power Snapdragon processor that would later fly aboard NASA’s Mars helicopter. At JPL, he led aerial mobility research and helped develop robotic systems for caves, subterranean networks, and other environments where maps, GPS, and reliable communication are not available.

That work eventually led to the DARPA Subterranean Challenge. Ali’s team sent legged, wheeled, and flying robots into unknown environments to map, explore, and coordinate missions with no prior information. The experience changed how he thought about autonomy. In the physical world, the unusual case quickly becomes the normal case.

That idea now sits at the center of FieldAI’s approach. Instead of asking a data-only model to learn every physical rule from raw camera and LiDAR input, FieldAI combines data-driven learning with physics and uncertainty quantification. The goal is to help a robot recognize when it is approaching the limits of what it knows, slow down, and make a safer decision.

Brian and Ali dig into what robot hallucinations look like when an AI model controls a physical machine. A wrong answer from a chatbot can be corrected. A wrong move from a robot around people, equipment, or an active jobsite carries a much higher cost.

They also examine why FieldAI runs its autonomy entirely on the robot without Wi-Fi, 5G, or a cloud connection. Ali argues that building safety into the architecture from day one allows robots to enter real customer workflows, collect useful data, improve, and create the deployment flywheel that robotics has struggled to start.

Construction is especially valuable because the environment changes by the hour. FieldAI is also deploying in energy, manufacturing, logistics, and urban operations, giving its models experience across very different conditions and tasks.

Finally, Ali explains why FieldAI is staying robot-agnostic while other physical AI companies move toward building full hardware stacks. Its software is already operating across 34 different robot embodiments, from multi-ton vehicles to quadrupeds and humanoids.

Connect with Ali Agha
https://www.linkedin.com/in/ali-agha-7aa5212a

Learn more about FieldAI
https://www.fieldai.com/

Learn more about FieldAI’s Field Foundation Models
https://www.fieldai.com/news/fieldai-announces-over-400m-in-funds-raised-to-advance-embodied-ai-at-scale

We’d love to hear from you.

Have thoughts or guest suggestions?

Reach us at podcast@automate.org

You can find the transcript and more episodes of Automated at automated.fm

Unlock full access to Automated and explore everything automation.
Subscribe today and leave a review on YouTube, Apple Podcasts, and Spotify. Also subscribe to the newsletter.
https://www.youtube.com/@automatedpodcast
https://podcasts.apple.com/us/podcast/automated-with-brian-heater/id1837762221
https://open.spotify.com/show/60olq6brlBEIJWggx2fMR6
https://www.automate.org/automation/automated-newsletter

You can also find us on:
LinkedIn https://www.linkedin.com/showcase/automated-podcast-by-a3/
Instagram https://www.instagram.com/automatedpod/

LinkedIn
https://www.linkedin.com/showcase/automated-podcast-by-a3/

Instagram
https://www.instagram.com/automatedpod/

Ali Agha on Robot Hallucinations, Physics, and Real-World AI

Automated with Brian Heater

Ali Agha on Robot Hallucinations, Physics, and Real-World AI